Experientia, 17: 72-73 (1961)
The anthocyanins of roses. Occurrence of peonin.
Harborne, JB and Corner, JJ

Although the garden rose contains a great range of colour varieties, only two anthocyanins have been so far been identified in the petals of cyanic forms. Cyanin (cyanidin 3:5-diglucoside) was isolated from Rosa gallica by WILLSTÄTTER and NOLAN in 19151and a pelargonidin 3:5-dimonoside (presumably the 3:5-diglucoside, pelargonin) was reported in the scarlet polyantha varieties 'Gloria Mundi,' 'Prince of Orange' and 'Paul Crampel' 2,3. The related flavonols, quercetin and kampferol were also known to occur in glycosidic form in rose petals. A third flavonol, myricetin, was recently described as occurring in about 20 Hybrid Tea varieties by SESHADRI et al.4. Since current work in this laboratory has shown that delphinidin and its methylated derivatives occur in association with myricetin in purple or mauve petals of a number of garden flowers5,6, a search for a delphinidin derivative among the anthocyanins of roses was undertaken. In particular, blooms of well established purple and mauve varieties as well as those of the latest and bluest breeding lines were examined.

No delphinidin was found and a re-examination of the rose varieties reported to contain some myricetin4 showed that only kampferol and quercetin were present. In the course of this Survey, however, a third major anthocyanin was discovered in Rosa rugosa and derived varieties, e.g. 'Roseraie de L'Hay'. The pinkish red petals of these plants contain cyanin and the new pigment, which was readily identified as peonin (Tab. I). Since peonin is rare and has only previously been found in quantity in peony blooms7, its presence in pink roses provides a valuable alternative source. Other new sources are the pink flowered garden geranium (Tab. I), a plant already known to contain pelargonin and malvin2, and dark red varieties of Lathyrus odoratus8.

Of the two previously known anthocyanins of roses cyanin is the most widely distributed, being present in all but two of the hundred or so varieties examined. The pelargonidin derivative, whose identity with pelargonin has now been confirmed, occurs in a number of scarlet varieties (e.g. 'Radar' and 'Will Scarlet') besides those already mentioned. Colour in the rose is therefore mainly due to pelargonin, cyanin or peonin or to mixtures of these pigments. Traces of the related 3-glucosides accompany these 3:5-diglucosides in some varieties. Purple or mauve colours are produced by co-pigmentation of cyanin; a fact which has been established by spectral measurements of aqueous acid extracts of the appropriate varieties (Tab. II).

  1. R. Willstätter and T. J. Nolan, Liebigs Ann. 408, 1 (1915).
  2. G. M. Robinson and R. Robinson, Biochem. J. 28, 1712 (1934).
  3. R. Scott-Moncrieff, J. Genet. 32, 117 (1936).
  4. S. R. Gupta, K. S. Pankajamani, and T. R. Seshadri, J. Sci. Ind. Res. B. (India) 16, 154 (1957). [Journal of Scientific & Industrial Research]
  5. J. B. Harborne, Biochem. J. 68, 12 P (1958).
  6. J. B. Harborne, Biochem. J. 74, 262 (1960).
  7. R. Willstätter and T. J. Nolan, Liebigs Ann. 408, 136 (1915).
  8. J. B. Harborne, Nature, Lond. 187, 240 (1960).